;----------------------------------------------------------------------- ; Copyright (C) 2001-2002 ; Associated Universities, Inc. Washington DC, USA. ; ; This program is free software; you can redistribute it and/or ; modify it under the terms of the GNU General Public License as ; published by the Free Software Foundation; either version 2 of ; the License, or (at your option) any later version. ; ; This program is distributed in the hope that it will be useful, ; but WITHOUT ANY WARRANTY; without even the implied warranty of ; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ; GNU General Public License for more details. ; ; You should have received a copy of the GNU General Public ; License along with this program; if not, write to the Free ; Software Foundation, Inc., 675 Massachusetts Ave, Cambridge, ; MA 02139, USA. ; ; Correspondence concerning AIPS should be addressed as follows: ; Internet email: aipsmail@nrao.edu. ; Postal address: AIPS Project Office ; National Radio Astronomy Observatory ; 520 Edgemont Road ; Charlottesville, VA 22903-2475 USA ----------------------------------------------------------------------- This file must be used to record any changes made to the Master version of AIPS kept in SOCORRO. An entry is: line 1: Date Program name Your name line 2: What was done (may span multiple lines) line n-1: Where has this change been moved (e.g. 15JUL94, nowhere) line n: Blank The details of line 1 are important since programs are used to find specified strings and do conversions on them. It MUST be typed as follows: 1 blank, 4 digits, a period, two blanks, then the Month, day, comma, year, some space, a title, some more space, your name. *********************************************************************** *********************************************************************** Changes while 31DEC02 is TST *********************************************************************** *********************************************************************** 10986. December 13-14, 2001 Update Eric Update: CHANGE.DOC, GETRLS, SETPAR, INSTEP1, MAKE.MNJ, update.pl, install.pl, AIPSPATH.SH and AIPSPATH.CSH to change 31DEC01 to 31DEC02. Changed COMRPL and COMLNK to update on BOB in the AOC when BOB is runningb as aipspgmr. Changed CDVER.SH and CDVER.CSH in all umpteen areas to define UPDATE to point at $TST/$ARCH/UPDATE except that UPHOME set right on LInux will cause it to point at $TST/UPDATE for mnj uses during putbck and the like. Moved NOWHERE - important not to go to NEW. 10987. December 17, 2001 DFTPL Eric The code that finds the I Stokes was really messed up! It counted the correlators that make up the I and then divided by that count. But the count was zeroed way outside the loop and the other parts were in the innermost part of the loop. And there was no test for zero divide. Moved to 31DEC01 this date. 10988. December 19, 2001 VLBACALA, VLBASUBA, VLBAFQS Amy The biggest change was made to VLBACALA where I changed the CPARM(6) (max deviation for the amplitudes) in the second run on SNSMO to 1.0. Since the max deviation is in real amplitudes (i.e. not fractions or percents) 0.02 was much too small. Changed VLBASUBA and VLBAFQS so that the default of CLINT is actually 1 minute as claimed in the help file. Previously it had been left to the default in INDXR which is 5 minutes. Moved to 31DEC01 this date. 10989. December 28, 2001 update.pl Eric Fixed to edit PAPC.INC into $INC/NOTST/$ARCH whether or not it started there. Previously it edited it in $INC! Moved to 31DEC01 this date.. 10990. December 29, 2001 CPASS.HLP Eric The help file lied about the auto normalization after averaging which was also done in CPARM(8)=0. Fixed that and added comments about normalizations. Moved to 31DEC01 this date. 10991. December 29, 2001 UVWAX Eric/Joe Lazio/Wendy Lane New task - like UVNOU but deletes data near 0 in u and/or near 0 in v and/or at high values of w. Moved to 31DEC01 this date. 10992. January 7, 2002 RMSD Leonia/Eric New task - computes the rms by self-consitent methods in a two-dimensional moving window through each plane of an image. The output is either the computed rms or the input image with blanking based on the computed rms. The latter, after transposition, should be an interesting input to XMOM. Moved to 31DEC01 this date. 10993. January 8, 2002 SETSP Eric Fixed error handling - a failure on one computer would cause it to fail on the next one too. Also it would not allow us to set a parameter to the value -1, quitting after a message with no warning when one tried. Moved to 31DEC01 this date. 10994. January 9, 2002 install.pl Eric Changed DADEVS handling to force the entry of a directory name only if $NET0/DADEVS.LIST does not exist. Added copy of files from $NET0/$HOST to $TST/$ARCH/TEMPLATE. Moved nowhere. 10995. January 9, 2002 TACOP Eric Changed to copy any kind of extension file. Changed the TACOP.FOR, TACOP.HLP and created $APLSUB/FILCOP.FOR. Itwas capable of trying to copy PL files! Now it will properly. Moved nowhere. 10996. January 9, 2002 IMAGR Eric Changed QCLEAN - apparently it was possible for a user to get the Clean windows messed up enough to confuse the I/O in the 2 routines that read only the window areas. Added more lines of code to defend against bad window parameters. I don't know that this will work but the problem has occurred only once that I know of. Moved nowhere. 10997. January 10, 2002 Plot tics Eric Full-sky plots like AITOFF did not work: DIRCOS The cutoff at 0.5 left off the edges in AIT - changed to 0.5000009D0. TICINC Did not reset the error code if there were some good and some bad positions - so it appeared that an error had occurred. CTICS Had some funny code applying only to the tick lengths for DOACROS true. It made the ticks too small so dropped it. Also put code in to not draw lines longer than 0.05 of an image when tracking along a coordinate. ITICS As CTICS TKTICS As CTICS. Moved nowhere. 10998. January 11, 2002 PASWRD Eric Corrected PASWORD to PASSWORD. Moved nowhere. 10999. January 12, 2002 RMSD Leonia The maximum size of the image plane increased to 8192. Moved nowhere 11000. January 28, 2002 BPASS Eric Added the option to normalize after the scan average of the data. The parameter BPASSPRM(5) has new values for this. Moved nowhere. 11001. January 29, 2002 FITLD Eric The file type BA is reserved in AIPS. FITLD was using this to reresent the unused possibility of bandpass tables. Changed BA to BT and changed DBAV.INC, BAINI.FOR, BAIO.FOR, BATAB.FOR to BT versions. Moved nowhere. 11002. February 1, 2002 WRTPROCS Amy A new package of procedures which write and read to and from DISK. They are: WRTDISK - Writes files to disk (with FITTP) using a standardized name. The user inputs the first and last catalog numbers to write to disk and WRTDISK loops through these numbers. The standardized names WRTDISK produces are designed to be read back into AIPS using READISK. READISK - Reads files from disk using standardized names. The user inputs the first and last file number to read from disk, and READISK loops through these files. READISK is designed to read in files that were written by WRTDISK. FITDISK - Writes files to disk using an automatically generated names bases on the AIPS file name. Moved nowhere. 11003. February 6, 2002 IMAGR Eric Changed: 1. The histogram now uses 16384 rather than 8192 levels. 2. The message following a filtering function used a FORMAT intended for another purpose. Added a suitable one. 3. When one forces a field and gets one with no components, the previous action was to remake all images. Changed to require that you force 5 bad fields before a remake is done. The count is reset only when a full remake is done. Moved nowhere. 11004. February 6, 2002 DBCON Eric Changed it to concatenate a list of file types from the two files. The list is CL, FG, TY, WX at present. It did do CL and FG before. Moved nowhere. 11005. February 11, 2002 SAD, JMFIT. IMFIT Eric Changed SAD, IMFIT, JMFIT to use the pointing position correctly for both the primary beam and the delay beam corrections. It used the image phase center for the former (!!!) and used a pixel offset for the latter. Changed the subroutine ELIPSQ to use DOUBLE PRECISION internally for all parameters since it was producing inaccurate answers for highly elliptical Gaussians. Moved to 31DEC01 patches page. 11006. February 11, 2002 $YPGNOT/CONFI.FOR Leonia CONFI has rescaled the size at Y direction to have the identical scale at both axis at the plot. It was done to see the ellipse edge as an ellipse. Otherwise it had been seen always as a circle. Previously the Ysize was rescaled. The ellipse always is elongated to N-S. So the rescaling of Ysize has brought to the loosing the top part of the ellipse. I have rescaled Xsize instead now. Moved nowhere. 11007. February 11, 2002 VLBAUTIL Amy Fixed one and added two procedures: VLBASUBS - Changed this procedure so that it checks if the input data file needs sorting before it sorts. Also, VLBASUBS used to not do anything if a subarray condition was not found, now it will run MSORT and INDXR as needed even if it does not need to run USUBA. VLBASRT - A new procedure that sorts (with MSORT) and indexes (with INDXR) a data file, if it is called for. In other words it checks if the file in in TB order and if a CL and NX table are present. VLBAMPCL - A new procedure which finds manual instrumental phase corrections with FRING. It allows the use of 2 scans in order to correct all antennas. Hopefully this will cover 99% of the cases. Moved nowhere. 11008. February 14, 2002 install.pl Pat Added support for CDROM installation, so that it doesn't do or say really dumb things. Fixed problem with fishing the compiler options out of the CCOPTS.SH, LDOPTS.SH and FDEFAULT.SH files, inserted some sanity checks for compiles working. Moved to 31DEC01 CD master this date. 11009. February 16, 2002 XAS.SHR, LIBR.DAT Eric Rearranged the logic of installing xas since DEC machines cannot handle the [ -f ... ] && structure. Added comments about -lncurses to $SYSLINUX/LIBR.DAT. Neither should cause anyone to change anything locally although there will be warnings to do so. Moved nowhere. 11010. February 16, 2002 WXINI, POINI Eric These routines used FNDCOL wrongly - namely they ignored its output and set the pointers as if all was well. For WX it was not because the column labels for the VLBA output WX table and those coded into WXINI were different. Reverted to the VLBA ones since I do not expect to be able to change them and added support for both. Moved nowhere. 11011. February 16, 2002 WETHR Eric New task to read in, smooth, and then plot the weather data or add to a flag table based on the weather data. Moved nowhere. 11012. February 18, 2002 SNPLT Amy SNPLT formerly only potted one stokes when DDLY (dispersive delay) was selected. This was fixed. Note however, that the DDLY in for both RR and LL will be identical. Moved nowhere. 11013. February 19, 2002 PRTAB Eric Added adverb NDIG as a switch on precision from 1PE10.3 to 1PE13.6 for singles and 1PD13.6 to 1PD17.10 for doubles when NDIG > 3. The 10.3 gets more columns across a page but the accuracy is limited. !PE13.6 is aboput the full limit on singles. Moved nowhere. 11014. February 19, 2002 IMAGR Eric The parameter IMAGRPRM(11)'s scaling was applied only to the average flux in the clean windows and not the peak. This error was put into the code when the selection of which field to clean was overhauled 2000-12-28. Moved to patch area for 31DEC01 this date. 11015. February 20, 2002 VLAPROCS, VLASUMM Amy Add VLASUMM to VLAPROCS. VLASUMM is identical to the VLBAUTIL procedure VLBASUMM. This procedure runs PRTAN on all antenna tables and then runs LISTR in SCAN mode. Moved nowhere. 11016. February 20, 2002 VLBASUMM.HLP Amy Corrected statement about how to make VLBASUMM available. on all antenna tables and then runs LISTR in SCAN mode. Moved nowhere. 11017. February 21, 2002 $APGNOT/RMSD Leonia It has been clear that the task RMSD is a very time consuming task especially for the big size of the image cube. To weaken the problem, the increments at X and Y are added as input parameters. The intermediate points are calculated by interpolation. The expected gain in time should be ~XINC*YINC Moved nowhere. 11018. February 21, 2002 BPASS Eric Changed to to make SOLINT infinite by default for single-source files. Previously it divided by zero when users thought scan average means something in single-source. Moved nowhere. 11019. February 25-26, 2002 ELIPSQ Eric This routine still does not always work right when it ought to be simple. Added code to scale the beam simply when CX=CY. Had a typo in that fix until 2/26. Moved last fix to AOC this date. 11020. February 26, 2002 CLIPM Eric A misspelled variable caused it to ignore the ANTENNAS and BASELINE specifications. Moved nowhere. 11021. February 26, 2002 DEFLG Eric Not properly tested: the number of polarizations was not reset to the 2 in use in the scratch file causing an I/O error. The time integration test was not done correctly causing it to integrate over the whole time and delete a lot of data. Moved nowhere. 11022. February 26, 2002 UVMOD Eric/Michael Bietenholz Misspelled variable caused it not to compute the RL, LR model correctly. Moved nowhere. 11023. February 28, 2002 IMAGR Eric Changed IMAGR.FOR to copy the input BOXFILE to the output OBOXFILE when both options are specified and when OBOXFILE does not pre-exist. Moved nowhere. 11024. March 1, 2002 CLCOR Amy The previous version of CLCOR did not handle the case of a single I stokes properly. If the STOKES input was blank, it would check to see if there was a single R or a single L stokes and if not then try to correct 2 stokes in the CL table, if there was only one stokes in the CL table, then the columns for the first stokes would be corrupted. This has been fixed to allow stokes I as well as adding the option to input STOKES 'I'. Moved nowhere. 11025. March 4, 2002 FITTP, FITAB Eric Added the official reference to FITS to the output headers, but blocked it from being copied from the HI file so that it appears only once. Moved nowhere. 11026. March 4, 2002 FACES Eric Comments in the help and Fortran were found to be nonsense left over from SETFC or wherever. No change to code. The same was true of the help file for CHKFC. Moved nowhere. 11027. March 6, 2002 TVPL Eric The scratch buffer was only 1280 and there are now laptops with TV windows wider than that. Changed to 2560. Moved nowhere. 11028. March 6, 2002 BOXES Eric New task - reads a BOXFILE and copies it, replacing the Clean boxes with ones taken from around the source in a source list. Fixed SETFC and FACES to read the source list and use all lines except those with a semi-colon in column 1. Fixed all source lists in $AIPSTARS and the ftp site to have semicolons in column 1 of all comment lines. The skip 31 lines algorithm was aging in a hurry. BOXES will allow source widths in the source list and make larger boxes if needed. Moved nowhere. 11029. March 7, 2002 VLBAUTIL Amy Changed several things in VLBAUTIL. VLBALOAD - Added merging of attached calibration tables (TAMRG). VLBALOAD now tests if there are TY and GC tables attached to the data and if more than 1 file was loaded and if so runs TAMRG for the TY, GC and PC tables. VLBAFIX - New procedure intended to replace VLBAFQS, VLBASUBS and VLBAFPOL. This procedure tests the data to see if it is in 'TB' order, if there are NX or CL tables present and if it is multifreqency and performs these steps: 1) If it is out of TB order then it runs MSORT 2) If the input SUBARRY > 1 it runs USUBA 3) If it is multifrequency it splits the data into single frequency files (with UVCOP) and the file is indexed with INDXR. 4) If it has only one stokes then VLBAFPOL is run. 5) If step 3 has not been performed and in the first NX or CL table are missing it runs INDXR. VLBASNPL - Added the option to plot dispersive delays (DDLY). VLBAMPCL - Fixed a bug, before it was only doing the final CLCAL on the source named in SOURCES. Moved nowhere. 11030. March 11, 2002 MFPRT Eric Added OPTYPE adverb with 'LIST' to make a source list in the format suitable for BOXES, SETFC, FACES. Moved nowhere, 11031. March 11, 2002 FLATN Eric Removed limit of 64 pointings and changed it to use the average value of NOISE for all pointings > 64. Moved nowhere. 11032. March 11, 2002 TVLOAD Eric This widely used subroutine also has buffers (1280) which are not big enough for Miller's and Greg's laptops. Changed to 2560 and relinked a bunch. BSCLN also had a 1280. Moved to Miller's machine. 11033. March 13, 2002 TVFLG, SPFLG Eric Changed them to leave DOCAT=-1 work files on disk when they die of unnatural causes (after the grid file is made anyway). Moved nowhere. 11034. March 13, 2002 UPDCVS Stephan, Eric The TEXT areas were not being updated - so new AIPSPUBL files and the changes to the AIPSTARS files were not propagated. Probably all files from TEXT areas will have to be cvs copied - an expensive proposition. Moved nowhere. 11035. March 13, 2002 TKERASE Eric New verb - erases the TEK screen. Changed AU9A, POPDAT, NEWPARMS, and TKERASE.HLP. Moved nowhere. 11036. March 19, 2002 ARRAY Leonia The count of rows included the number of columns (array axis 1) in both read and write of whole arrays. This was not used before MAPBM so was not previously corrected. Moved nowhere. 11037. March 19, 2002 ABOUT files Eric SHOPH calls for update on: ZZANALYSIS.HLP CALIBRAT.HLP EDITING.HLP FITS.HLP IMAGE-UT.HLP IMAGING.HLP MODELING.HLP PLOT.HLP PROCEDUR.HLP TABLE.HLP TASK.HLP UTILITY.HLP UV.HLP VERB.HLP VLBI.HLP Moved nowhere. 11038. March 20, 2002 $APGOOP/MAPBM Leonia The original version of MAPBM created the average maps for all 4 stokes averaging all selected antennas and all selected reference antennas. The new version of MAPBM allows to create the stokes maps for each of the selected antenna list separately. Each output map is the cube with the third axis dimension - 4 (one plate for each Stokes parameter). So the AIPS task KNTR can plot all polarizations at one page. The cube of the antenna number '#' has the class 'ANT_#'. The four maps of the cube can be normalized to 1(one) to simplify selection of the levels at KNTR. The value of the scale factors can be found as keywords using the aips verb 'imh' Moved nowhere. 11039. March 20, 2002 CVEL, BPASS Eric The hairbrained scheme to carry around which antenna is which for shifting using the 0.001 position in the baseline code will never work. Changed it to use antenna 2 unless VLBA. This is an error only if antennas are renumbered for non-VLBA, VLB data. Which antenna is which matters not for small arrays like the ATCA and VLA. Moved nowhere. 11040. March 22, 2002 $YPGNOT/CONFI.FOR Leonia The multiplication of the array beam by the antenna primary added as an option. This option allows optimization of the array side lobes taking into account the antenna primary beam. Just now there is only one type of the primary beam corresponding to the circle dish of the given diameter with homogeneous illumination. Moved nowhere. 11041. March 23, 2002 WRTPROCS Eric Changed INDISK to OLDVOL in READISK. INDISK is a special adverb with a host-specific limit that may not match the old host's limit. Dropped OUTNAME and OUTCLASS as dangerous. Made minor edits to proc and help file. Moved nowhere. 11042. March 24, 2002 VLAPROCS Eric A procedure was included that referenced 2 VLBAUTIL procedures that were not themselves defined. Moved nowhere. 11043. March 24, 2002 POSSM Eric Corrected the plotting and labeling of the multi-IF plots (APARM(9) > 0) and changed it to offer multi-polarization as well. It will now loop over the requested polarizations. Moved nowhere. 11044. March 24, 2002 NEWPOS Eric Removed the message about L, M out of range. Error handling is now good enough that we don't need the message - esp when it will be generate 1000's of times. Moved nowhere. 11045. March 26, 2002 POSSM Eric Fixed a couple of typos that caused it to mishandle bandpass plots, especially those for LL-only data sets. Moved nowhere. 11046. March 27, 2002 CookBook Eric Putbck the $DOCTXT/LaCook,tex and $AIPSPUBL/COOK.PS for Chapter what 0 Update Table of contents 3 FITS disk procedures 4 BPASS normalizations, TVFLG DPARM(5) option, WETHR, FILLM CL table actions 5 BOXES, new BOXFILE options and freer format 6 PRTAB (NDIG), WETHR, TKERASE, TACOP (PL files etc) 7 COMB (SUMM), RMSD, SAD changes, IMFIT, MFPRT option for BOXES et al 8 RMSD 9 VLBAFIX, VLBASUBS, VLBAMPCL; POSSM plot options, remarks about BPASS changed 13 Update all ABOUT files C Overhauled by Amy to emphasize the VLBA procs including new ones D New high-frequency appendix by Lorant F Rename D. I Update index. Moved to the web site this date. 11047. March 27, 2002 install.pl Eric Removed one last message about ssh from install.pl. Moved nowhere. 11048. March 31, 2002 SG2RUN Eric Changed STORES, K2TEXT, and AU2A to implement a new verb to copy the K array to a text file suitable for later use as a RUN file (e.g., on a computer of another architecture). Also POPSDAT, NEWPARMS, and a help file. Moved nowhere. 11049. April 9, 2002 IMAGR Eric Messing with multi-resolution Clean: 1. Changed Q1GRD, QDFT, QGRIDA, QTAPER to compute the argument to the EXP function and only do it when > -14. This greatly speeds up things with substantial taper. 2. Changed IMAGR.FOR to record ROBUST specially in the HI file so that a value of 0 will be displayed. Also changed to pass NGAUSS (as NUMRES) in the Clean object. 3. Changed IMAGR.HLP to do a MAXPIXEL control over resolutions along with the other knobs. 3. Changed QCLEAN to pick up NUMRES, to set the field weights before the first TV display selection, to increase the TV size correctly when interpolating the image, and, if the number of fields/resolution > 2, to sum the histogram only over the fields of the particular resolution. It reports the settings of FACTOR, GAIN and MAXPIXEL for multi-resolution cleans and adjusts maxpixel in IMAGRPRM(16). Moved nowhere. 11050. April 9, 2002 MNJ Eric Changed $UPDUNIX/UPDCONFIG to do $SYSLOCAL/UPDCONFIG silently. It was sending a special cron "error" message on this now standard message. Changed MAKE.MNJ to determine the CVS before building the $SYSLOCAL/UPDCONFIG and to include cvscmd=$CVS in that file. Also changed it so that aipsmgr@nrao.edu only gets error messages not success messages. Moved nowhere. 11051. April 9, 2002 XAS.SHR Eric Dropped -m486 from the Makefile. It is an obsolete option and probably did not help anything and it hurts when trying to compile on a Mac or other non-X86 cpu. Moved nowhere. 11052. April 10, 2002 FITAB Eric FITAB read the full 14 random parameter names and used all that were not blank and not REMOVED, WEIGHT, or SCALE. This ignores NRPARM which should be a limit on the list. Moved nowhere. 11053. April 10, 2002 CALIB Eric Put in header tests to block polarization and IF averaging when there are not suitable things to average. Moved nowhere. 11054. April 15, 2002 WRTPROCS/FITDISK Amy FITDISK now accepts a ':' after the logical variable representing the output disk, which is stated in OUTFILE. This is how WRTDISK always worked, I just wanted to make them consistent since they are part of the same "package". I also added WRTPROCS.HLP, which just states what procedures are part of WRTPROCS and give a brief description. Moved nowhere. 11055. April 16, 2002 XAS.SHR Eric Jonathan Quick pointed out that /bin/sh rather than the elaborate UNSHR will work on XAS.SHR if the first of each pair of "magic" strings is enclosed in quotes. Now do we get rid of UNSHR or not? Moved nowhere. 11056. April 23, 2002 POSSM Eric Fixed an addressing error affecting BP plots when fewer than all spectral channels are plotted. Moved nowhere. 11057. April 23, 2002 FILLM, WETHR Eric Pulled the opacity correction out of Bryan's new FILLM as $APLSUB/OPACTY and $APLSUB/SATPRS subroutines. Fixed WETHR to use them as well - required adding a new adverb to WETHR. Changes to FILLM by Bryan Butler and me: 1. Fixed up history writing to tell what happened. 2. Changed the help file to be more descriptive of changes in FILLM. 3. Documented what was done to the data in terms of the opacity and gain curve effect on the CL table, both in the message file, and in the attached history table. 4. Added a "seasonal" opacity model, and now allow the user to have any weighted combination of the surface weather and seasonal opacities. The default is a straight average (weights both 0.5). 5. Changed, marginally, the coefficients for opacity calculations, based on more TIP data. 6. Changed, marginally, the calculation of saturation vapor pressure (made it the right one, instead of barry's old approximation [which is pretty good, but as long as i have a better one, i think it should be used]). 7. Fixed a glaring bug in the setting of the gain curve coefficients. Turns out that i (or Eric) had never changed the indexing of the 'antenna 0' entry to be MXANT+1 instead of 0 in FLMGN. so, the 'antenna 0' coefficients were still written into the 0'th array entry, which was a boundary violation, and it was actually writing over other variables. sheesh... Moved nowhere. 11058. April 26, 2002 APGPS Eric The routine that accessed the FQ table declared the IF frequency to be REAL, but the table routines handle it as DOUBLE PRECISION. Fixed the declaration - I don't know that this task is used but that routine is handy for the INDXR rewrite upcoming. Moved nowhere. 11059. April 29, 2002 $QPGNOT/UVCON Leonia The new option to multiply the model by the antenna primary beam has been added. The primary beam can be chosen as: circular dish with the flat illumination circular dish with illumination 10db down at the dish edge circular dish with illumination 15db down at the dish edge The primary beam applies at the both model given as clean component and as an image. Moved nowhere. 11060. April 29, 2002 AIPSUPD Eric On failure it would display <= 60 lines of the failure which is often too few. Changed to 125 lines. Moved nowhere. 11061. April 30, 2002 INDXR Eric Changed TABLEUTIL to provide access to the WX subroutines. Changed INDXR to compute the CL table for weather and antenna gains using the routines like in FILLM. The results are not the same exactly due to differences in the zenith angle which are due to the use of a circular earth model in GETANT and SOUELV. The differences are rather less than the uncertainties in the antenna gains and opacity models. Moved nowhere. 11062. April 30, 2002 REGRD Eric Changed the help to document new CPARM default - all zero will use the input reference pixel value (transformed) for the output. Before one had to force cparm(1-3) and (5-7) each out of range to get that behavior. Changed the code to take advantage of LOCNUM (avoids lots of COPYs and SETLOCs) and to work on data cubes - it did one plane only before. Moved nowhere. 11063. April 30, 2002 XAS.SHR Eric The latest versions of Linux RedHat 7.2 XWindows seem to generate an extra screen event. This triggered XAS to rethink its center pixel prematurely. It got set by a YWINDO call but the screen changed size a bit later - after the event and so after the centers requested were changed to accomodate a window larger than the one about to be used. Moved the code that reconsiders the centers to just before the place where those values are really used. This seems to fix the problem. This change is serious enough for all MNJ sites to install it. Remember UNSHR is not needed - just sh XAS.SHR will do. Changed instructions in UPDCONTROL for this. Moved nowhere. 11064. May 3, 2002 Adverb pass back Eric Added the capability to pass adverbs back to aips from tasks. Changed: AU1A New verb OUTPUTS lists output adverbs AU2 Notice codes for get back adverbs and get them on GO AU2A TGET only gets adverbs that GO puts. PTPARM New: puts parms back into the TD file for AIPS to read HELPS Add OUTPUTS pseudoverb POPSDAT Add OUTPUTS pseudoverb NEWPARMS Add OUTPUTS pseudoverb OUTPUTS Help file IMEAN Return the PIXAVG and PISTD adverbs IMEAN Help file mentions this and has new codes ($) There is a code in column 10 of the inputs to indicate an adverb's uses: ' ' GO, '*' GO and TELL, '?' TELL only '&' GO, TELL, GET, '%' TELL and GET only, '$' GO and GET '@' GET only. Moved nowhere. 11065. May 7, 2002 VLBAUTIL Amy I made all of VLBAUTIL more compact by putting multiple statements on each line. Also, changed VLBALOAD, VLBAFIX and VLBAFQS: VLBALOAD - fixed a couple of minor bugs: changed query about whether TAMRG should be run so that it only asks if NCOUNT <> 1 rather than > 1 since NCOUNT=0 can mean multiple files are loaded; also added TGET VLBALOAD at the end of the procedure. VLBAFIX - Substantial rewrite of the frequency copying portion. Added ability split S/X and 90cm/50cm simultaneous observations. In these cases the S/X and/or 90/50 are stored as different IFs in the same FREQID. However, VLBAFIX will not handle 90/50/20cm when observed simultaneously. VLBAFQS - Same change as VLBAFIX. Moved nowhere. 11066. May 7, 2002 VPLOT Eric Claimed that the number of adverbs was n + MAXFLD -> it is really n + MAXAFL the number of fields in AIPS adverbs. Moved nowhere. 11067. May 8, 2002 Number of fields Eric The Clean box arrays were the limiting factor in memory usage limiting the number of fields. With multi-resolution clean forcing the number of fields = NFIELD * NGAUSS the limit of 512 looked inadequate. So changed: PUVD.INC Raised MAXFLD to 4096. PCLN.INC Changed parameters so that the max # Clean boxes per field will be min [ 4096, (64*4096)/N ] where N = total number fields = NFIELD * NGAUSS. TVUTIL Made Clean box array use dynamic memory. QCLEAN Changed field number of digits in Formats. IMAGR Made Clean box array use dynamic memory, new format output file names. WFCLN Made Clean box array use dynamic memory, new format output file names. SCIMG Made Clean box array use dynamic memory, new format output file names SETGDS Changed file class naming ICL001-ICL999 then IC1000-IC4096 etc. BLING File naming has not worked since MAXFLD was 16. GPHAS Updated file naming. Changed formats in CALIB, FRING, MX. Changed HELP files for limits: IMAGR, SCIMG, WFCLN, BOXFILE, BLCAL, BLING, BPASS, CALIB, CLPLT, CPASS, FLATN, FRING, GPHAS, IBLED, KRING, LPCAL, PCAL, UVADC, UVCON, UVSUB, VPLOT. Moved nowhere. 11068. May 9, 2002 IMAGR Eric IMAGR, SCIMG, and WFCLN had the wrong type declared for WIN and so, when WIN was set from FLDSIZE, it ended up with bad values going into the Clean - unless they were reset with the TV or BOXFILE. Sigh - all my tests had a boxfile. Moved nowhere. 11069. May 10, 2002 Misc Eric Fixed SAD.HLP to talk about CPARM(i) instead of the old CUTOFF. Fixed LaCook4 to avoid mention of BCHAN and ECHAN relative to BPASS; they are not used by BPASS. Moved nowhere. 11070. May 10, 2002 UVDEC Eric New task: copies every n'th spectral channel. Moved nowhere. 11071. May 13, 2002 Misc Eric Changed: VPLOT Fixed number adverbs from 330+MAXFLD to 330+MAXAFL which then exposed the fact that it is 331. IMAGR Fixed QCLEAN - the menu item to SELECT FIELD became SELECT FLD when the field numbers added a digit. LaCookC BPARM=0,2 to get phase plots in Sect 3.4 #8. Moved nowhere. 11072. May 13, 2002 APCAL Amy Changed APCAL so that it can handle TY tables with more than 4096 entries. This involved increasing MXVAL from 4096 to 8192 and MXBUFF from 130000 to 260000. Moved nowhere. 11073. May 15, 2002 VLBAFIX Amy Fixed VLBAFIX so it works with files that observed S or X but not S/X. Moved nowhere. 11074. May 15, 2002 various help files Amy FLAGVER.HLP - used to say FLAGVER=0 means do not apply flag table, this is not true for many tasks. Now says it is different for different tasks. FITLD.HLP, INFILE.HLP, CALIBRAT.HLP - remove reference to ANCAL. Moved nowhere. 11075. May 15, 2002 install.pl Eric Changed the code that copies files to $AIPS_ROOT from $SYSUNIX. This copy checked dates but there are ways for that to fail so I made it copy the files all the time. If there are restarts this blows away the .old files but so what. Also added $mytype = "TST" and the MNJ code will be done only if $mytype is /TST/. When we get a new NEW this changes to NEW and the MNJ stuff is skipped as it should be for NEW. Moved nowhere. 11076. May 16, 2002 FITLD Eric Penticton needs the VIS scaled by SCLVIS but not by integration time, EVN needs the vis scaled not at all, and the VLBA needs the vis scaled by SCLVIS * int time * WAIT. All 3 now get the weights scaled by INT time. Moved nowhere. 11077. May 21, 2002 $QPGNOT/UVCON Leonia UVCON can use different (4) coordinate systems for the input antenna position including the mixture of the coordinate system. Sometimes it is required to have all antennas at the same coordinate system. The new version allows to reformat the input file to the OUTFILE recalculating all antenna positions to the equatorial coordinate system. UVCON can continue the rest of the job or exit having finished reformatting under control of BPARM(10) Moved nowhere. 11078. June 6, 2002 $QPGNOT/UVCON Leonia The multiplication of the model by the primary beam can be carried out now with the given shift of the primary beam. This means that UVCON having repeated with the different shifts is ready for simulation of the mosaic observation. The help file is updated. Moved nowhere. 11079. June 11, 2002 POSSM Amy There were 2 display bugs. 1) STOKES 'LR' never plotted. 2) if the number of baselines that existed in a single scan was < NPLOTS, then all the scans would plot on top of each other until the number of the baselines was >= NPLOTS. Both of these bugs were small typos introduced when POSSM was revised to allow multi-polarizations to be plotted. Moved nowhere. 11080. June 14, 2002 BLAPP Walter Alef/Eric Put Walter's changes to support the output of MK4IN - a task not distributed with AIPS since it is used at the MKIV correlators to translate to AIPS uv FITS and uses a lot of the Haystack software package. Moved from MPI this date. 11081. June 14, 2002 IBLED Eric Changed the name of the menu item to select a 2nd IF for an IF-ratio display. Also changed the initial value for flag all channels to true. Moved nowhere. 11082. June 15, 2002 CVEL Eric Chris Phillips has pointed out that the shift antenna is ANT1 for all arrays except the EVN corrrelator - if there has been no reordering of course. The planned way to store this was never honored and would probably not work (lost in floating-point inaccuracies) anyway so let's hope that the order of antennas does not get switched outside of the VLBA where it does not matter. The EVN knows of this and will be careful - but the others??? Moved nowhere. 11083. June 18, 2002 SPLAT Eric When averaging to a smaller number of spectral channels, SPLAT had a possible data addressing error when the user-specified BCHAN to ECHAN range was not an exact multiple of CHANNEL. The error would arise in the second IF and was accompanied by a warning message that some channels were being dropped rather then being included in the averaging. Moved to patch area 20 June. 11084. June 20, 2002 FITLD Eric The routines that read in tables assume that there are data in, for example, the PH table only for allowed FREQIDs. Apparently, the VLBA correlator can write bad data - messing up which FREQID is which - so some defense against unknown FREQIDs was added for some of the routines. The recovery plan is not guaranteed to be correct and it was installed only for GN, PH, and TS (GC, PC, TY) tables. In so doing, an error in examining the source table was found. The file skipping when the requested source(s) were not present worked only on the first tape file. Now it will happen for all files that do not include the requested sources. Moved nowhere. 11085. June 20, 2002 SAD Help Eric CUTOFF was still referenced although it is now CPARM(i) for more than one i. Moved nowhere. 11086. June 20, 2002 VLBAUTIL Amy VLBACRPL - changed the default STOKES to 'I' rather than leaving it blank, which plots 'FULL' (i.e. RR LL LR RL). VLBAFIX - fixed looping when VLBAFPOL is called. Moved nowhere. 11087. June 24, 2002 VLBACALA Amy Removed REFANT input. Set REFANT = 0 in procedure. It was an input to SNSMO that doesn't seem to be used since SNSMO is just smoothing amplitudes. Moved nowhere. 11088. June 27, 2002 IMAGR Eric Changed QCLEAN.FOR to report the final cleaned flux even if there is only 1 field. Moved nowhere. 11089. June 28, 2002 Misc Eric FLGIT The report of previously flagged used to require a correction but in the rearranging of counters the correction no longer belongs. It was producing negative numbers when VPOL clipping is significant. SETFC Help file required correction/clarification. FACES Help file required correction/clarification. BOXES Help file required correction/clarification. Moved nowhere. 11090. June 28, 2002 $YPGNOT/FRPLT.FOR Leonia FRPLT which makes plots of the fringe rate spectrum, has been had limited options to make a plot. For example, FRPLT has been able to plot only one baseline at one run; there has not been the option to repeat plots for each next SOLINT interval or each next scan.... The new version of FRPLT can: 1. plot many plots at one page (different baselines, IFs, polarizations....), 2. select the baselines using input parameters ANTENNAS and BASELINE by the standard way, 3. repeat plots for each next SOLINT interval or each next scan. 4. plot several IF's and/or polarizations together as though one long spectrum. 5. plot Amplitude & Phase (default) or Amplitude, or Phase or ..... 6. plot different combinations of polarizations. Moved nowhere. 11091. July 2, 2002 disk files Eric Changed $APLUNIX/ZPATH.C to allow a disk file name with no "logical:" in it. Changed $APLGEN/ZFULLN.FOR to trap a name that begins with "/" and treat that as an absolute path name. Changed $APLGEN/ZTPOPN.FOR to call ZFULLN with default $FITS for FITS disk files - the change to C I/O in ZTPOPD caused it to skip applying any default directory. Changed input name length to 256 in $APLUNIX/ZTPOPD.C. Changed all name lengths from 128 to 256 in the following. The max path name in Unix is 256 and some people have found the 128 limit onerous: $AIPSUB/AU5C.FOR AUC.FOR $APLSUB/LPOPEN.FOR OFMDIR.FOR WRIBOX.FOR $APLUNIX/ZDIR.FOR ZTTOPN.FOR ZPATH.C ZTPOPD.C $APLGEN/ZPATH.FOR ZFULLN.FOR ZIVSOP.FOR ZLASIO.FOR $APLGEN/ZLPCLS.FOR ZLWIO.FOR ZLPOPN.FOR ZMOUNT.FOR $APLGEN/ZMYVER.FOR ZTOPEN.FOR ZTPOPD.FOR ZTPOPN.FOR $APLGEN/ZTXMAT.FOR ZTXOP2.FOR ZVERSN.FOR ZTPOPR.FOR $APLGEN/ZRMTAP.FOR $AIPGUNIX/ZSTRTA.FOR ZSTRTB.FOR ZSTRTD.FOR $AIPNOT/TPMON.FOR $AIPPGM/GRIPR.FOR $APGNOT/VLBIN.FOR $APGOOP/HF2SV.FOR CL2HF.FOR $QPGOOP/OMFIT.FOR WFCLN.FOR Moved nowhere. 11092. July 2, 2002 BPASS Eric BPASS needs the same change as CVEL. The shift antenna is 2 for the EVN and 1 (or really none for the VLBA) for the rest. Moved nowhere. 11093. July 2, 2002 BOXES Eric For BOXFILES without F or C cards get max field number from the Clean Boxes already present. Moved nowhere. 11094. July 3, 2002 Help files Eric Clarified the wording describing NCOMP in help files for BLCAL, BLING, BPASS, CALIB, CLPLT, CPASS, FRING, GPHAS, IBLED, KRING, LPCAL, NCOMP, PCAL, UVADC, UVCON, UVSUB, and VPLOT. Reworded helps for INFILE, IN2FILE, OUTFILE, OUTPRINT, OFMFILE, BOXFILE, and OBOXFILE to account for the new file name handling. Created a help file for IMAGRPRM. Moved nowhere. 11095. July 6, 2002 $QPGNOT/UVCON.FOR Leonia The geodetic latitude, longitude coordinate system is converted now by the more precise routine GEOXYZ.FOR, taken from SCHED, written by Craig Walker. The help file is updated. Moved nowhere. 11096. July 4, 2002 UVFND Eric The number of channels to print and check was computed using the calibration output header, not its input. So it would only do 1 channel for CHANNEL > 5. Moved nowhere. 11097. July 6, 2002 MFPRT Eric MFPRT's LIST option was not doing the correct coordinate computation. Also it tested IMSIZE widths even when not set. Moved nowhere. 11098. July 6, 2002 Help for SETFC, FACES, BOXES Eric "Image name" should have been "UV dataset name" Moved nowhere. 11099. July 6, 2002 SNPLT Eric Increased the work area size by a factor of 6. Moved nowhere. 11100. July 7, 2002 GRDCRM Eric difmap insists on writing a CC file with 7 columns even though it is doing point models. This confused GRDCRM into thinking that things do not match since they were not type Gaussian. Changed the test to look for type 0 also (point) and compare the widths there too. Moved nowhere. 11101. July 8, 2002 CONFI.FOR .HLP Leonia The height of the array center above the geoid is added to the output file header, when the output file is at the UVCON format. The help file is updated. Moved nowhere. 11102. July 8, 2002 install.pl Eric Changed default to start INSTEP2 and 4 rather than not to start them. Moved nowhere. 11103. July 11, 2002 CONVL.HLP Leonia It has not been clear from the help file what scale is carried out under control of FACTOR if the input image is measured in Jy/pixel. One short paragraph has been added to clarify it. Moved nowhere. 11104. July 12, 2002 VLBAUTIL Amy VLBAFIX had a bug where if it had to fun VLBAFPOL after it seperated frequencies it did it on the original file rather than the UVCOPied file. Fixed this as well as changed VLBAFPOL so if it is called by VLBAFIX after splitting frequencies it uses a sensible OUTCLASS. Moved nowhere. 11105. July 12, 2002 $APGNOT/PCCOR.FOR Leonia PCCOR has not recalculated the time for SN table entry(using the the previous row time instead) if the current PC row has not to be corrected (because for example the antenna did not have PC measurement at the calibrator time interval). Moved nowhere. 11106. July 14, 2002 WIPER Eric New task: Does a UVPLT-like plot into memory, then interactive editing on TV image of the memory array, then flags to an FG table. Moved nowhere. 11107. July 15, 2002 WIPER Eric Added ABORT, FLAG FAT, and UNFLAG FAT options. The latter 2 are a fast mode with a cursor of radius BPARM(9) plot pixels. Created HLPWIPER.HLP for on-line assistance. Fixed a failure to handle large IMSIZEs properly and a bad message. Moved nowhere. 11108. July 15, 2002 START_AIPS Eric Changed the order to define things with AIPSPATH and AIPSASSN before setting the socket info. The latter needs REHEX which may not be in the path until the former are done. Moved nowhere. 11109. July 15, 2002 SPLAT Eric SPLAT copied tables from BIF to EIF even if it averaged the IFs. Changed to copy BIF to BIF in that case. Moved nowhere. 11110. July 16, 2002 WIPER Eric Added WIPER SET and WIPER SIZE interactive methods of setting the fat wiper size. Made the fast modes require a button hit to get started. Moved nowhere. 11111. July 16, 2002 SPLAT Eric If DOCAL true no CL table should be copied - even CL 1 has stuff in it these days. INDXR will make a new one. Moved nowhere. 11112. July 16, 2002 VLBAUTIL(VLBAFIX) Amy Fixed yet another bug in VLBAFIX. Formerly thought that a file had been observed in SX if there was a large jump in frequency between FREQIDs and one of the frequencies observed was S or X. Moved nowhere. 11113. July 16, 2002 UVCON Leonia/Eric UVCON has not written the position at the pointing direction. The absence of the position at the pointing direction creates the problem for the task VTESS for example. Moved nowhere. 11114. July 17, 2002 WIPER Eric Fixed bugs in counting - it counted fully flagged input records as flagged - and in writing flags - it got the source number wrong (-1 not something useful). Also added a message about the actual Stokes used and plotted. Fixed help DOALL range and inputs description Moved nowhere. 11115. July 17, 2002 RENAME Eric Changed AU7.FOR - RENAME checked for a duplicate name on OUTSEQ=0 only within the input type. This could allow it to have duplicate names, one for UV and one for MA. But many tasks and verbs do not take INTYPE as an argument and so get confused. Moved nowhere. 11116. July 17, 2002 MAKE.MNJ Eric Changed the output do_daily. file to export $PATH if it changed it (or put a commented export path otherwise) and to invoke LOGIN.SH from $AIPS_ROOT rather than the . which is unlikely to have such a file. Moved nowhere. 11117. July 17, 2002 UVDIF Eric Changed the conversion of data so that weights are more readily accessible. Then added option to check for differences in weights and to choose to ignore flagging. Moved nowhere. 11118. July 18, 2002 QUACK Eric Added option TAIL to flag all sources for APARM(2) minutes after each scan. This covers the case in which the array thinks it has moved on to the next source but some antennas have not yet gotten the message and so think themselves still on source (the last one...). Moved nowhere. 11119. July 22, 2002 UVCON Leonia The shift of the simulated UV data relatively given center (RA=0, DEC=APARM(3)) has been added. So now UVCON can simulate the mosaic observation with the tangent point of each pointing coincided with the pointing direction. I have not removed the previous 'interpretation of mosaic when the tangent point of each pointing stayed at the same direction - center of mosaic map. These two options are under control of input parameters. UVCON has had three versions of simulated primary beam. The all three are related with the illumination of the dish. Now I have added the 4th version of the PB which is not related a dish but is determined by the Gaussian beam parameters: BMAJ, BMIN, BPA. This option simplifies working with other AIPS task which use the Gaussian beam. Moved nowhere. 11120. July 22, 2002 CORFQ Eric This task for some reason thought that it would not work on compressed data. It prepared a message to that effect but neglected to call MSGWRT before dying. It works fine on compressed anyway. Moved nowhere. 11121. July 22, 2002 FILLM Eric Changed FILLM to allow the reading of disk files. This requires a new $APLUNIX/ZDKMID.C routine and a new FLDKIO routine inside FILLM.FOR. Moved nowhere. 11122. July 23, 2002 CORFQ Eric Apparently there are compilers out there that do not initialize the memory to zero - good luck to them!!!! Changed NCOL to be initialized to the correct value, although 0 is also correct for pre-existing tables. Moved nowhere. 11123. July 23, 2002 FILLM Eric FILLM checked that the file count was exactly NCOUNT but had places where it could exceed NCOUNT bypassing the check. Changed it to bail out where it will print the tape file # message if the count is too high and to check in the main spot for >= NCOUNT. Moved nowhere. 11124. July 23, 2002 MOMNT Eric The help file lied about the support size for Gaussians but the HI file got it right. Moved nowhere. 11125. July 25, 2002 CLCAL Amy Added input DOBLANK which controls whether blanked values in an SN table are replaced if smoothing is requested. If DOBLANK >= 0 then blanked values are replaced by the smoothed values (calculated only from non-blanked solutions). Also, in this situation, SMOTYPE='FULL' is forced since not to do so would introduce bad data into the interpolation. This only changes the behavior of CLCAL if INTERPOL = 'BOX' or 'MFW' and DOBLANK >= 0. Moved nowhere. 11126. July 25, 2002 FILLM Eric Found an error in EOF handling - if an INIT call produced an EOF then an INIT call was done followed (on success only) by an OPCODE (init remember) call. Two INITs in a row is a bad idea esp for disk files but it at least produces a wrong file count with tapes. Improved the history writing and removed a serious error. In an attempt to avoid math aborts, I added a test on the scaling exponent but I made the test way too limiting and so weak sources were 1000's of times too strong on output. Moved to AOC $LOAD areas immediately. 11127. July 26, 2002 BLOAT Eric This old task could not read compressed data and did the addressing of the uv data incorrectly. It also used the number of Stokes to limit the number of channels to copy. Moved a test version to VLA Linux $LOAD area. 11128. July 26, 2002 WIPER Eric If one requested only one subarray, then flags were written for subarray 1 even if you were plotting subarray 2. Fixed logic. Moved nowhere. 11129. August 7, 2002 CPASS Eric Fixed it to pay attention to the error returns from the fitting programs (usually due to input error) and some other routines as well (task used IRET when IERR was returned from the main subroutine). Also put a test on the tolerance to avoid one means by which the thing failed (the test inside has a fixed number - 10 - while the task sets a relative tolerance to compare with it). Added usage warning to the help file. Moved nowhere. 11130. August 7, 2002 IMAGR Eric Made it more tolerant of loading 0 pixels - it will go on trying (ignoring that field) until 5 fields load 0 pixels. Also fixed it to separate histograms on resolutions even if there is only one field at a per resolution. Dropped a dumb message. Moved nowhere. 11131. August 9, 2002 install.pl Eric The INSTEP2 and FILAIP questions were case sensitive in their answers. Moved nowhere. 11132. August 9, 2002 FILLM Eric The channel 0 data had a bandwidth (and weights) based on 0.75 of the total bandwidth (e.g. 8 * chan_wid) when only 7 channels gets used and only 5 are are summed locally. Fixed the bandwidth computation. It also had the ability to skip a channel 0 sample because the channel 0 from the on-line system was pure zero when the spectral samples are not pure zero. While such data are bad, the discrepancy in vis numbers is confusing. Moved nowhere. 11133. August 9, 2002 UVDIF Eric Added an option to allow the program to go ahead even when parts of the header disagree. Fixed error on testing whether data are good. Moved nowhere. 11134. August 9, 2002 AIPS Eric KLEENEX invokes KILSRV which tries to kill TV displays even if there are none. Added an IF statement. Moved nowhere. 11135. August 13, 2002 Perl Eric I guess I am going to have to learn Perl. Errors in the Perl in DADEVS.PL caused the start up for machines like RM251A-B to include a great many computers (why did no one tell me????) and in install.pl caused it to do the opposite of the users request on whether or not to start INSTEP2/4. Moved nowhere. 11136. August 13, 2002 CookBook Eric Changed chapters: 0 For changes to T of C etc 3 SG2RUN, full path names in file names 4 FILLM disk files, INDXR does WX corrs, WIPER 5 4096 fields limit, WIPER 6 WIPER, IMEAN returns adverbs 7 IMEAN returns adverbs 9 QUACK tail, CLCAL uses DOBLANK, xrfe, FRPLT changed 12 SG2RUN C REFANT not in VLBACALA I For index changes. Moved nowhere. 11137. September 3, 2002 TPDEVS.SH Eric The grep "^$HOST" would get wrong entries on AAA if there was one host named AAA and another named AAAB for example. A simple change to "^$HOST " should do the trick. Moved nowhere. 11138. September 4, 2002 ZCREA2.C Eric Changed error messages to show errno. Moved nowhere. 11139 September 9, 2002 FITLD Leonia The input parameter GR for the subroutine JPRECS has the wrong value .FALSE. As a result FITLD has calculated the apparent position of sources with small error. I changed it for .TRUE. Moved nowhere. 11140. September 9, 2002 SAD Eric The position angle mess in SAD continues to bite us. It carefully noted that PA's in SAD are different by 90 - unfortunately after they were converted to the normal CCW from N. The error bars in X and Y were the most affected, the CC file written might also be bad. Moved nowhere. 11141. September 12, 2002 NFILES limit Eric Changed NFILES limit from 999 to 9999 in FITLD and IMLOD help files. Moved nowhere. 11142. September 16, 2002 BPASS Eric Egregious error caused it to miss the first record of every SOLINT integration period except the first. Moved nowhere. 11143 September 17, 2002 IMAGR Eric The parameter FGAUS was mishandled so that the 0-width set of fields did not get a flux cutoff. Also added FGAUS to the HI file. Fixed a curious bug in QCLEAN: having chosen a field and re-imaged it, the program did not notice that its maximum was less than the minflux for that field. In the histogram computation, however, it skipped the field and then had to deal with ahistogram containing no points. The result was that all points were loaded to the AP and 1 component taken before quitting for the flux level. Fixed to have an error (no points found sort) from the hitogram routine. Moved nowhere. 11144. September 24, 2002 *TESS Leonia/Eric Changed to allow up to 4096 fields (pointings) in the tessalation routines DFIL.INC Changed to allow 4096 rather than 128 BMSHP Add message on BMAJ going in to FITBM VTTELL Corrected number of parameters, fixed to handle fields > 64 VTESS Fixed formats, drop VMBLKD, set beam size and error for fields > 64 from #64 UTESS Fixed formats, drop VMBLKD, set beam size and error for fields > 64 from #64 LTESS Fixed formats, drop VMBLKD, set beam size and error for fields > 64 from #64 STESS Fixed formats, drop VMBLKD, set beam size and error for fields > 64 from #64 IMERG Got rid of VMBLKD. VMBLKD Removed. Moved nowhere. 11145. September 25, 2002 CLCOR Eric Changed the time interval for computing parallactic angle from 1 sec to .02 sec. Fixed unused variables. Moved nowhere. 11146. September 27, 2002 PATGN Eric Added RADI - a radial polynomial function and POLY an X and Y polynomial function. Moved nowhere. 11147. September 27, 2002 install.pl Eric AP size 20 Mbytes was recommended on machines > 70 Mbytes which is not enough for the OS these days. Changed to recommend 80 Mbytes over 400 Mbytes and 20 over 140 Mbytes of RAM. These are more realistic. Moved nowhere. 11148. September 30, 2002 install.pl Eric Added code to make a FITS directory if one is not present. Moved nowhere. 11149. September 30, 2002 LIBR.DAT Eric For Linux replaced -ltermcap with -lncurses for $AIPPGM. The latter works on all systems while the former only works for now on RedHat. Moved nowhere. 11150. September 30, 2002 SEARCH Eric The SEARCH proc attempts to find which version of the files is the most recent for compiling and linking purposes. It makes a list of the possible file names and then looks to see if there is more than 1. The previous method could get the wrong answer when there were other decimal points in the file name besides the main one - i.e. a path such as /.net/home/aipsetc Changed the method of decidiung that there are two and added code so that the order of prog.f and prog.FOR in the list was not assumed (new versions of ls ignore case some). Moved nowhere. 11151. September 30, 2002 RMSD Leonia OPCODE at RMSD has had the two values 'rms ', 'blan'. There has not been a default for OPCODE. I added the default ('rms ') for OPCODE. Moved nowhere. 11152. October 1, 2002 VTESS.HLP Leonia Small changes near NOISE and PBSIZE have been done. Moved nowhere. 11153. October 1, 2002 Large images Eric FFT had a declaration that limited the image size to 4096. TVCPS had two formats that limited the image size to 9999 in black and white and 3333 in color. Raised both limits to MAXIMG or more. Moved nowhere. 11154. October 2, 2002 OOP Eric CLASSIO.INC has a hollerith buffer of class names that could hold only about 640. Changed it to hold 5000. Moved nowhere. 11155. October 4, 2002 ZDCHIN Eric Changed DEVTAB(37-39) to indicate tapes. This may propogate eventually. Moved nowhere. 11156. October 8, 2002 UVMOD Eric Corrected handling of GPOS for phase shifts. They were taken as specifying L and M where they are supposed to specify increments on coordinates. Calls to XYSHFT and SHISIN are needed to turn the increments into terms used for phase shifting. Moved nowhere. 11157. October 21, 2002 POSSM Eric Had some problems left over from when it did only 1 IF and pol or multiple IF but 1 polarization in the plots. Should stop better when at end and should honor NPLOTS better. Moved nowhere. 11158. October 22, 2002 CLCAL Eric Removed illegal Fortran: WRITE (MSGTXT,1020),XDOBLK moved nowhere. 11159. October 23, 2002 SHISIN Eric The Z term in the correction was computed as the sqrt(1-dx**2-dy**2) which is accurate in small-medium angle cases. But the < 00 sqrt is needed sometimes for large angles. Replaced the s qrt with the correct trigonometric formula. Moved nowhere. 11160. October 23, 2002 scripts Eric Various scripts used ei1 and ni1 for procs but they call each other causing some redefinition confusion. Made unique names for each in TVDEVS.SH, STARTPMON, START_TVSERVERS, UNIXSERVERS, and XASERVERS. Corrected grammar in TVDEVS.SH - a $USEUNIXSOCK was not surrounded by "'s and so caused failure when undefined. START_AIPS nslookup is obsolete on Linux. Added a case $ARCH. MAKE.MNJ needed to ask for the start time if missing first and then apply corrections and tests. Added correction for yyyymmdd.* where the * is incomplete. Moved nowhere. 11161. October 30, 2002 BPASS Eric Added option to rescale the BP solution to correct for a calibrator spectral index. Also changed POPSDAT.HLP, NEWPARMS.001 to add adverb SPECINDX and a new help file. Moved nowhere. 11162. October 30, 2002 XMOM Eric Added option to scale the cutoff values by the primary beam. This allows XMOM to be run on the output of PBCOR which takes the frequency dependence of the primary beam properly into account. PBCOR will not work on the moment omages since they do not have an identifiable frequency. Moved nowhere. 11163. November 1, 2002 VLBAMPCL Amy Added ECHAN and BCHAN to VLBAMPCL since this is written with spectral line observations in mind and they may want to leave out bad end channels. Moved nowhere. 11164. November 5, 2002 IMAGR Eric There was an error copying the Clean boxes from the 0-resolution fields to the extended resolution fields that could get things all wrong, even messing up the Clean boxes of the zero resolution. Fixed IMAGR.FOR. Moved nowhere. 11165. November 5, 2002 LOCIT Eric New task to fit antenna locations and new run file procedue BASFIT.001 to aid analysts in running CALIB, LOCIT, and LWPLA. Moved nowhere. 11166. November 6, 2002 NCHAV Eric New adverb - put in place of NPOINTS in the help files for COMAP_MX, IMAGR, MAPIT_MX, MAPPR, MX, SPLIT, WFCLN and in the default handling in AIPS AU2A.FOR. Changed POPSDAT.HLP, NEWPARMS.001, and created NCHAV.HLP. Also CookBook chapters 4, 8, 9, and C. Fixed CROSSPOL reference in 9. Moved nowhere. 11167. November 11, 2002 UVCON.FOR, .HLP Leonia The following three modifications have been added: 1. The simulation of the random phase added to each antenna. This option is requested by Rick Perley to simulate the random phase added to an antenna through the very far side lobes when the array observes at one direction but RFI comes from the very far away direction. 2. UVW have been calculated for the source direction RA=0, DEC=APARM(3)not paying attention on the recently implemented RASHIFT, DECSHIFT. Corrected. 3. The new Eric's subroutine SHISIN has been implemented to address better the case of the wide field of the model. Moved nowhere. 11168. November 12, 2002 TVLOAD Eric The routine that scales the data for a TVLOAD, ISCALE had a typo so that negative square root was not done. Moved nowhere. 11169. November 13, 2002 PCAL.HLP, LPCAL.HLP, SPCAL.HLP Amy Added paragraph(s) to help files in order to clarify which should be used in each case. Moved nowhere. 11170. November 15, 2002 install.pl Eric Create $AIPS_ROOT/OFM directory. Make $RUNFIL, $FITS, $OFMFIL group writable always. Moved nowhere. 11171. November 15, 2002 IMAGR Eric Fixed bug in selecting the first image to display in multi-resolution Clean. Moved nowhere. 11172. November 25, 2002 VLBAPIPE.001, VLBAPIPE.HLP Lorant Added the VLBA data calibration procedure into the system Moved nowhere. 11173. November 27, 2002 OFMFILEs Eric Added several OFMs - basically the 6 TVPHLAME colors and the 4 step functions of TVPSEUDO (button C). Moved nowhere. 11174. December 2, 2002 Plotting Eric Added several new plot command types: line type (1 labeling, 2 contour, 3 pol vectors, 4 stars et al.), dark vector, init for color, and changed code number for init for grey since added PIXRANGE to it. Changed: DGPH.INC Added GPHRNG (pixranges), GPHFUN (functype), GPHDOD (do dark), changed GPHTVC and GPHTVG to arrays of 3 and 8, added GPHLTY (current line type), GPHCON(16) (channel mask off/on), GPHCUT (cut value between dark and bright vectors) G3COLR New: write a 3-color "grey-scale" - code 11. GCHAR Init and use needed/current graphics channel. GCINIT New: init a channel if not already done GDVEC New: write a dark vector - code 10 GETOFM New: read OFM from TV/file, init the TV GFINIS Change to multi-channel writing of image catalog GINIT Init new parameters. GPHTVG(i) = i if GRCHAN=0, else GPHTVG(i)=GRCHAN, GPHTVG(8) = 8 for dark, drop channel inits. GINITC New - init for 3 color graphics - code 12 GINITG Changed code to 13, add FUNCTYPE, PIXRANGE out. Use these in image catalog - previously the plot could not be used in CURVAL. GLTYPE New - write line type command - code 9. GRAYPX Use GCINIT to init channel, hold # 3000 -> 5000. GSCALE New - scales a floating buffer of image values to an integer buffer for plot incl GPHRNG and GPHFUN types. GVEC Changed to init/use current graphics channel. Hold off at 5000 not 3000. Moved nowhere. 11175. December 1-2, 2002 Plot to device Eric POPSDAT Add adverbs: DODARK, DOCOLOR, PLCOLORS, DARKLINE NEWPARMS Add adverbs: DODARK, DOCOLOR, PLCOLORS, DARKLINE Also new help files for these four. TKPL Ignore new opcodes, advance beyond them, draw dark vectors as bright. TXPL Ignore new opcodes, advance beyond them, draw dark vectors as bright. TVPL HELP: change description of GRCHAN, add DODARK. TVPL Make GRCHAN and IMCHAN arrays, new selection of GRCHAN as GINIT, multi-channel image catalog at end, init channels when needed, use graphics function and real pixrange in image catalog, handle new 3 color plot type. LWPLA HELP: add DODARK, DOCOLOR, PLCOLORS, OFMFILE. LWPLA New adverbs. Support new functions. The 4 line types are allowed separate colors in both bright and dark forms. Exterior characters and the background are also allowed colors. DOCOLOR must be true to have PLCOLORS(3,10) be used to set these colors. The task can color grey-scale images with an OFM controlled by OFMFILE from a text file or the TV. Moved nowhere. 11176. December 2, 2002 Plot tasks Eric Changed plot tasks: DCNT.INC Define buffers for contouring/grey scale. AU2A DEFAULT for DARKLINE AU5B Change call sequences to PNTPLT and STARPL. AU8A EXTLIST - upgrade for new GREYS and PCNTR. COMLAB Add call to GLTYPE (1). CONDRW New call sequence to allow for background gray scale to cause vectors to be dark. Drop TELL. CTICS New call sequence for PLOTWO, drop DGPH.INC. LABINI Add 0.5 characters to left CLAB1 Move vertical axis label in by 0.5 characters CLAB2 Move vertical axis label in by 0.5 characters GTICNT New: Counts label characters for step wedge PLOTWO New call sequence - dark or bright vectors? PNTPLT New call sequence - dark or bright vectors? - pass on to PLOTWO STARPL New call sequnce to pass in the background gray image. Reads background image and plots with dark or bright symbols. GREYS HELP: add FUNCTYPE to allow functions and DODARK to allow choosing dark or bright graphics vectors when DOTV > 0. Added OFMFILE. GREYS Use DCNT.INC, add FUNCTYPE, DODARK. Compute step functions and plot values with PIXRANGE and GSCALE routine. Changed STARPL and CONDRW call sequences to allow dark vectors with grey image. Added code to read in an OFM, init the TV properly, and make a 3-color pseudo-color plot. New plot type number (27). PCNTR HELP: add DODARK, 4th image, PIXRANGE, FUNCTYPE, DOWEDGE, OFMFILE, TVCHAN. PCNTR Use DCNT.INC, add DODARK and other adverbs. Use GLTYPE, changed call sequences to STARPL, CONDRW. Introduce the full set of grey-scale plotting with color, dark vectors over the top, etc. using image 4 for gray scale. New plot type number (28). KNTR HELP: add FUNCTYPE, OFMFILE, DODARK, DARKLINE, DOCOLOR, Images 2 and 4, DOVECT and polarization adverbs. KNTR Add coloring to grey-scales, add calls to GLTYPE, change call sequence to STARPL, revise contouring to allow dark contours. Display true color RGB images when DOCOLOR true. Add polarization vectors. CCNTR Use DCNT.INC, GLTYPE, new call sequence for STARPL and PLOTWO. CNTR Use DCNT.INC, GLTYPE, new call sequence for STARPL LOCIT Changed call sequence of PNTPLT (no dark here) PLOTR Changed call sequence of PNTPLT (no dark here) Moved nowhere. 11177. December 2 & 4, 2002 RGB tasks Eric Changed TVHUI to fix bugs in the handling of XINC > 1 and to have the output image ignore BLC and TRC as well as XINC and YINC. Added SQ transfer function. Also changed help file and HLPTVHUI help file. Found bugs in TVRGB: Due to an addressing error, it did not weight the red and blue equally although it should. Also it ignored the direction of the axis 3 coordinate and so red and blue had nothing to do with red-shifted or blue-shifted. Fixed these - added experimental options for doing the summing. Moved nowhere. 11178. December 4, 2002 BPASS Eric Changed help file to point out that DIVCH0 is false for autocorrelation only data. Add warning message to BPASS when BPASSP(10) is zero in this case. Moved nowhere. 11179. December 4, 2002 plot tasks Eric KNTR and PCNTR failed to allow room for displaying the contour levels unless polarization vectors were also plotted. KNTR also had trouble dealing with multiple panels including polarization vectors. Moved nowhere. 11180. December 4, 2002 VLBAPIPE.HLP Lorant Rewrote most of the help file with Amy's input to make it more understandable for the general user. Moved nowhere. 11181. December 4, 2002 GEOMs Eric Raised the buffer size in GEOM from 20000 to 600000. Raised the buffer size in LGEOM, HGEOM, and PGEOM from 700000 to 2400000. This still is not enough for really big images. Is it worth recoding to do in squares? Moved nowhere. 11182. December 5, 2002 PRTTP Eric Changed the code and help to allow VLA archive files from disk including a full directory's (= 1 tape = 2 Gbytes) worth of VLA data. Added to FILLM help file as well - revealing where the VLA archive is at the AOC among other things. Moved nowhere. 11183. December 5, 2002 VLBAPIPE.001 VLBAPIPE.HLP Lorant Added a DOUVCOMP option as well as a PL-file report at the end and removed message supressions after suggestions by Craig, Amy. Moved nowhere. 11184. December 6, 2002 START_AIPS Eric The default for REMOTE_ROOT should be AIPS_ROOT not /AIPS. Moved nowhere. 11185. December 6, 2002 LOCIT Eric Called DFILL when it should have called RFILL. Made address exception when optimised. Added calls to GLTYPE. Moved nowhere. 11186. December 9, 2002 TV buffers, XAS Eric When one has two monitors the TV buffer gets twice as big - too big for some of our buffers. Changed to have all TV buffers = MAXIMG which should always be too big. Added PMAD.INC to DTVS.INC, DGPH.INC, TVUTIL.FOR BSCLN, TVHUI, TVRGB, IBLED, IMLHS, SPFLG, TVFLG, TVDIC, WIPER Added DCNT.INC to OFMCON, OFMMOD, TVLOAD Fixed typo in PLOTR as well. XAS had numerous places which limited the image size to 2048 or perhaps even less. These should now allow 16384+16 bytes in I/O etc. Moved nowhere. 11187. December 9-13, 2002 Plot tasks Eric Fixed LOCIT to aoid plotting an antenna with no data. Added GLTYPE calls to: BPERR BSTST IMEAN IMRMS IMVIM ISPEC PROFL SERCH SL2PL TAPLT UVHGM PLAXES PLPNT PLLINE BPLOT CLPLT CONFI DFTPL ELINT FGPLT FRMAP FRPLT GAL LOCIT PBEAM PFPL1 PFPL2 PFPL3 PLCUB PLOTR PLROW POSSM SNPLT UVPLT VPLOT WETHR and fix up character counts. Other changes: SLBINI Add extra 0.5 character to left like LABINI LABINI Reduce CH(1) by 0.333 GINITL Make CHOUT(i) at least 0.833 PLGRY Change call seq to pass work buffer - which was only 4096 previously LWPLA Correct drawing of vertical strings - correct handling of old plots with grey scales. Add option to convert to CMYK rather than RGB. Incl help. PROFL Made it less sensitive to string lengths, fixed scaling to use full plot area. IMVIM Allowed more bins - up to 512,512. LOCIT Use different colors for symbols in crowded plots. Initialize variables better and reset warning error returns. UVPLT Add DOWEIGHT adverb to weight the binning sums and use BPARM(8) < 0 to allow the individual points as well as the bin averages to be plotted. They are different line types and so can be different colors. Moved nowhere. 11188. December 12, 2002 FUDGE Eric Added APARM(1) option to scale amplitudes of all data. Moved nowhere. 11189. December 16, 2002 Speed parameter Eric Changed FILAI2 and FILAIP to set the speed parameter onto disk. They ignored it previously. Also made 0 -> 10.0. Changed SETSP to display the value before asking whether to change it when doing all machines with individual values. Moved nowhere. 11190. December 16, 2002 procs Eric Changed FDEFAULT, CCOPTS, LDOPTS in $SYSLINUX to allow for the Intel compiler. Moved nowhere. *********************************************************************** *********************************************************************** Changes while 31DEC02 is NEW *********************************************************************** *********************************************************************** 11192. December 16, 2002 install.pl Eric Corrected the handling of the MNJ so that the date will be set by default to the tar ball date - 1 not December of last year. Moved from 31DEC03 this date. 11193. December 16, 2002 plots Eric CCNTR and CNTR had an excess call to GLTYPE ahead of GINITL which confuses LWPLA. Moved from 31DEC03 this date. 11194. December 17, 2002 LWPLA Eric Added code to prevent writing commands ahead of the first correct line in the PostScript output file (as CNTR caused until yesterday). Issues warnings, does not write bad lines, and continues. Moved from 31DEC03 this date. 11195. December 17, 2002 LOCIT, BASFIT Eric Changed BASFIT.001 to offer DOCOLOR option. Added info to BASFIT help file. Fixed bugs in LOCIT centering phases when some are blanked and printing phases when some are blanked. Moved from 31DEC03 this date. 11196. December 17, 2002 SUL procs Eric Changed FDEFAULT, CCOPTS, LDOPTS to allow g77 as well as SUNWspro compilers in $SYSSUL. WARNING: the g77 are untested. Moved from 31DEC03 this date. 11197. December 23, 2002 Plotting Eric LWPLA The defense against premature plot commands did not take into account the possibility of appending to a pre-existing file. The defense then stopped any appending. Added a second place for FIRST=.FALSE. KNTR Needed to manipulate the choice of LOCNUM for STARPL. Otherwise LTYPEs other than 3 could not plot stars. Moved from 31DEC03 this date. 11198. December 23, 2002 TBIN Eric The program did not handle a missing HI file properly. Added code to create one if there was none and to avoid catastrophic errors in there is an HI problem. Moved from 31DEC03 this date. 11199. December 23, 2002 POSSM Eric Had the ability to call GFINIS without having called GINIT. This wrote to the unopened TV and hung both POSSM and AIPS. Moved from 31DEC03 this date. 11200. December 24, 2002 LWPLA Eric Fixed so it will not repeat the setrgbcolor commands when not needed. Also fixed wrong format for setgrey. Moved from 31DEC03 this date. 11201. December 24, 2002 install.pl Eric The question about skipping the tar ball unpack had the wrong code for the default - requiring a Y or y to be typed to skip. Moved fix (not full file) from 31DEC03. 11202???